strisunshinewentingwang/pytorch

Pull Requests Commits

[optim][adagrad] group tensors in foreach to maximize perf

janeyx99 committed 2 years ago

67a1880c

Rename Makefile_dashboard to Makefile (#92584)

ezyang committed 2 years ago

a20c678c

Do not specialize int/float with dynamic=True (#92570)

ezyang committed 2 years ago

90024436

[dtensor] disable gpu tests in op db first (#92611)

wanchaol committed 2 years ago

0bc875ac

Fix/modernize dynamo docs (#92572)

wconstab committed 2 years ago

a2b8e891

Register sccache epilogue before starting sccache (#92587)

huydhn committed 2 years ago

ce43fc58

Reenable mobilevit_s in CI, seems to pass (#92585)

ezyang committed 2 years ago

44e52ea5

vmap support for torch.linalg.vander (#91749)

jazzysoggy committed 2 years ago

b6cfd622

[optim][radam] fix eps discrepancy for foreach (#92551)

janeyx99 committed 2 years ago

3ba5eae7

Run CI in a new environment (#92378)

malfet committed 2 years ago

97f34e36

[MPS] Fix index_select scalar input with multiple indices (#91064)

qqaatw committed 2 years ago

ccbdf495

Revert "[vmap] unsafe_split : batching rule and OpInfo (#92291)"

pytorchmergebot committed 2 years ago

827e22ec

[primTorch] Remove prims.to_dtype (#92380)

peterbell10 committed 2 years ago

a9f44628

[BE] Get rid of `future` (#92596)

malfet committed 2 years ago

1906eaf2

[reland] Improve hooks ordering behavior (#92559)

soulitzer committed 2 years ago

1bc60c6b

[vmap] unsafe_split : batching rule and OpInfo (#92291)

kshitij12345 committed 2 years ago

0510ae59

Follow up comments of PR #91531 (#92359)

zhaojuanmao committed 2 years ago

0a404fdd

Fix `ShardedTensorMetadata.tensor_properties` for Python 3.11 (#91795)

ConnorBaker committed 2 years ago

20665235

[threaded_pg] fix the comments of MultiThreadTestCase (#92373)

wanchaol committed 2 years ago

06d54b40

[dtensor] delete lagging op db and update op db tests (#92290)

wanchaol committed 2 years ago

997de441

Improve `bsr @ strided` performance in `baddmm` for `bfloat16/half` with Triton kernels. (#88078)

nikitaved committed 2 years ago

8383b5c4

some fixes to get symbolic shapes working through inductor (#92320)

Chillee committed 2 years ago

4f4b62e4

Fix key error formatting and move exc code to exc.py (#92593)

mlazos committed 2 years ago

cac217c8

Make run_dynamic_ci_skips_only.sh more generic (#92581)

ezyang committed 2 years ago

ba682057

[CI] move parallelnative to periodic (experimental) (#92567)

janeyx99 committed 2 years ago

2a7a859d

Remove temporary export skip hack (#92160)

voznesenskym committed 2 years ago

28cb3141

fix promote_constants with ExpandView (#92403)

Natalia Gimelshein committed 2 years ago

ef258642

When nopython=True, Dynamo can't allow graph breaks. (#90970)

wschin committed 2 years ago

bdbd3ed3

Guard on at::Tensor device index (#91779)

voznesenskym committed 2 years ago

eb39d990

[CI] valgrind 3.16.1->3.20.0 (#92552)

malfet committed 2 years ago

388d79cc

Older